OpenSRE is an open-source framework for building AI agents that investigate and resolve production incidents using operational data and tools. It is for site reliability engineers who want customizable incident-response workflows, training, and evaluation on their own infrastructure.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add Tracer-Cloud/opensre --skill c-delegating-github-ci-repairsgit clone --depth 1 https://github.com/Tracer-Cloud/opensreWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/tracer-cloud/opensre/c-delegating-github-ci-repairs)<a href="https://agentmods.dev/skills/tracer-cloud/opensre/c-delegating-github-ci-repairs"><img src="https://agentmods.dev/badge/skills/tracer-cloud/opensre/c-delegating-github-ci-repairs/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/tracer-cloud/opensre/c-delegating-github-ci-repairs"><img src="https://agentmods.dev/badge/skills/tracer-cloud/opensre/c-delegating-github-ci-repairs.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00036 | $0.01877 |
| Opus 5.5 | $0.00014 | $0.00751 |
| Sonnet 5.5 | $0.00007 | $0.00375 |
| Haiku 4.5 | $0.00004 | $0.00188 |
Grade A, and why
delegating-github-ci-repairs scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 181 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Delegate a remote CI repair
This runs one bounded repair that finishes on the gateway without the shell. It does not set up continuous repository monitoring.
Roles
Orchestrator (local shell):
- Checks the gateway is ready, picks the target, delegates, verifies and reports.
Executor (gateway):
- Calls
run_ci_repair_demoonce and reports that result back to the shell.
Plan
Use update_plan to create the live plan from the workflow headings below. Mark a step in_progress or completed in the same response as that step's tool call. A response that only calls update_plan is not progress.
Inside the interactive shell:
- Prepare the hosted gateway with check_hosted_gateway.
- Probe GitHub access on the gateway with one ask_hosted_gateway prompt.
Inside the interactive shell:
- Confirm the target and approval in one ask_user_choice question.
- Delegate execution with ask_hosted_gateway and retain its prompt ID.
Remote Gateway agent:
- Call
run_ci_repair_demoonce for the approved owner and repo, without loading a skill. - Respond with that tool's outcome as Markdown. Nothing on GitHub is deleted; the demo repository is kept.
Inside the interactive shell:
- Verify the outcome by re-reading the delegated prompt ID (verifies: true).
- Show the remote outcome and evidence as Markdown.
- Offer the next step or blocker resolution with ask_user_choice.
Success criteria
The workflow succeeds only when:
- The hosted gateway is healthy with active GitHub permissions.
- A seed demo repository is configured with a failing PR
- The CI failure is detected remotely.
- The remote repair loop fixes the failure without intervention from the local shell.
Workflow Notes
Prepare the hosted gateway
Shell only:
- Run
check_hosted_gateway(). If it is not running, callstart_hosted_gateway()once and check again until it is ready. - Do not send a test prompt. The probe below is the first prompt, and its answer proves the gateway takes prompts.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago Changed · +6 lines ea25625796b7
- 12d ago Changed · -11 lines 187f26e5fa20
- 13d ago Changed · +13 lines bfef292eb664
- 14d ago Changed 99c0a7dff146
- 16d ago Changed · +93 lines · -10 tokens per session a98de57e745d
- 17d ago Changed · -21 lines · -55 tokens per session 309532d11fd4
- 22d ago Changed · +66 lines · +61 tokens per session 2873215a9365
- 28d ago Changed · +10 lines 29d688c4b34e
- 1mo ago Changed · -4 lines aee4b68e1b9b
- 1mo ago First seen · 29 lines · 40 tokens per session scan A 87801e75aa06
delegating-github-ci-repairs is a skill published in the GitHub repository Tracer-Cloud/opensre (11,969 stars, last pushed today), licensed Apache-2.0. It adds 36 tokens to every session and 1,877 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-15.
Other skills, from other repositories
grafana-expert
Expert-level Grafana dashboards, visualization, data sources, alerting, and production operations. Use when the user mentions dashboards, visualization, monitoring, observability, or alerting, or when the task involves Grafana Architecture, Installation on Kubernetes, Data Sources, or Dashboard JSON.
host_bash
A read-only shell skill for examining files and running diagnostic commands in a restricted environment. A shell is a text interface for operating-system commands.
host_restart_service
A tool for restarting approved systemd services, which are background programs managed by Linux's service manager.
monitoring-expert
Expert-level monitoring and observability with Prometheus, Grafana, logging, and alerting. Use when the user mentions observability, Prometheus, Grafana, logging, metrics, or alerting, or when the task involves The Three Pillars of Observability, Monitoring Fundamentals, Prometheus Configuration, or Alert Rules.
datadog
Full-stack observability with Datadog APM, logs, metrics, synthetics, and RUM. Use when implementing monitoring, tracing, alerting, or cost optimization for production systems.
grafana
Build operable Grafana dashboards with template variables, provisioning-as-code, and unified alerts wired to Prometheus/Loki; avoid high-cardinality panels agents love to ship.